The setup note to write before you launch any campaign

by Harry Aloysius

In short 

• One page, written before launch, naming what the campaign is for and what would count as an answer. It takes about forty minutes and it is the cheapest insurance available on a paid budget. Its real function is to make the decisions while nobody is emotionally attached to the outcome.

• Written after launch, it is a report. Written before, it is the thing that stops the arguing.

Most campaigns launch with the settings decided and nothing else. Budget, audience, creative, all specified. What success looks like, when it gets judged, and who decides: all left to be worked out later, under pressure, by people with different assumptions.

The setup note fixes that. It is one page, written before anything spends, and it is the single highest-return forty minutes in paid advertising.

Why write it before rather than after?

Because after launch you are no longer a neutral party.

Once money is moving and numbers are arriving, every judgement is contaminated. A campaign that looks bad on day three feels like it needs rescuing. One that looks good feels like it needs more budget. Neither feeling is information.

Decisions made in advance are made by someone with no stake in the answer you, before you knew. That is the whole mechanism, and it is why the note has to be written first to be worth anything.

The secondary benefit is that writing it surfaces disagreements early. Half the time the note gets written and it turns out the client thought the campaign was for sales and the team thought it was for awareness. Finding that out on the page costs nothing. Finding it out in the review meeting costs the campaign.

What should you decide before you launch a campaign?

Nine lines. It fits on one page and it should.

Line	What it answers
The one sentence	What this campaign is trying to make happen
The outcome	The single event that counts as success, defined precisely
The target	What that outcome should cost, and where the number came from
The floor	The result at which we stop and rethink
The stopping rule	Spend threshold, metric, number — for individual ads
The review date	When we look, and what we will not touch before then
What we are testing	The one variable that differs between variants
The destination	Which page, and what it asks for
Who decides	The named person who calls it

Nothing about targeting settings, bid strategies or placements. Those belong in the account, not in the note. The note is about judgement, not configuration.

The one sentence

Write what the campaign is trying to make happen, in a sentence a person outside marketing would understand.

Good: "Get enquiries from D2C brands doing over two crore, for the retention audit."

Bad: "Drive awareness and engagement for the Q3 push."

The test is whether the sentence names a specific person doing a specific thing. If it names a marketing activity instead, the campaign does not yet have a purpose, it has a slot in a calendar.

The outcome

Define the event so precisely that two people counting would get the same number.

"Leads" is not precise. Is a newsletter signup a lead? Is a form fill from a student? Is a duplicate?

Precise: "A completed enquiry form from a business, excluding students and job applicants, counted once per email address."

This matters because the exclusions are where the arguments happen in month two. Writing them down now means the reporting cannot quietly redefine success later, which it will otherwise do, in the direction of whatever happened.

The target and the floor

Two numbers, and the floor is the one people skip.

The target is what the outcome should cost. Say where the number came from: past campaigns, what a customer is worth, or a stated assumption. "We think ₹X because a customer is worth ₹Y and we close about one in four" is a legitimate basis. "₹X" with no reasoning is a number someone will argue with later.

The floor is the result at which you stop and rethink rather than optimise. Optimising has a lower bound, if the cost is several times the target, the problem is not the bid strategy, and continuing to adjust settings is a way of not facing that.

Without a floor, campaigns get nursed indefinitely, because there is always one more thing to try.

The stopping rule

Three parts: a spend threshold, a metric, and a number.

"Any single ad that spends twice the target cost per enquiry without producing one is turned off."

This is the rule for individual ads inside the campaign, as distinct from the floor, which is for the campaign as a whole. It gets its own treatment in when to kill an ad, and the rule that decides it.

Write it in the note so that turning something off is executing an agreed rule rather than admitting a failure. That distinction sounds like semantics and is not, it is the difference between a decision that happens on time and one that happens three weeks late.

The review date

Name the date, and name what will not be touched before it.

"First review on the 21st. Nothing changes before then except stopping ads under the rule above."

The most common way budget gets wasted is not bad targeting. It is a campaign edited so often it never settles, because every edit restarts the platform's optimisation. Writing the date down converts "leave it alone" from a nervous instinct into an agreed position you can point at when someone asks why nothing has been changed yet.

One useful addition: name what you would consider an emergency. Something visibly broken wrong link, wrong creative, obvious overspend, is worth fixing immediately. "Performance looks disappointing" is not an emergency and should be named as not being one.

What we are testing

One variable. Write down which.

If the variants differ in image, headline and audience at once, the result is unattributable. Whichever wins, you have learned that this particular combination beat that particular combination, which is not transferable to anything.

The note should say: "Variants differ only in the problem named in the headline. Everything else held constant."

That also makes the write-up trivial later, because the question the test answers is already written down.

Who decides

One named person, not a group.

Campaign decisions made by committee drift, because nobody owns the moment where something gets stopped. Name the person who calls it. It does not have to be the most senior person in the room; it has to be one person.

Where a client is involved, this line also settles what they are deciding versus being told. Both are fine. Ambiguity between them is not.

What does the note change in practice?

Three things, in rough order of value.

Arguments become short. When the result comes in, the question is not "was this good" but "did it meet the target we wrote down." That conversation takes minutes rather than a meeting.

Bad campaigns die on time. The floor and the stopping rule remove the need for anyone to be the person who admits it failed.

Good campaigns get properly credited. A campaign that beats a target written in advance is much more persuasive evidence than one declared a success afterwards against no stated standard. This matters more than it sounds, particularly when next quarter's budget is being decided.

What to do this week

Write the note for a campaign that is already running. It is less useful retroactively, but it will immediately show you which of the nine lines nobody ever decided.

Add the floor to it. Most people can state a target and have never named the point at which they stop optimising and rethink. That number is the one that saves money.

Send it to whoever else touches the account. A rule one person holds privately is a preference. Written and shared, it is a rule.

The thing to remember

The note is not planning documentation. It is a set of decisions made by the only version of you who has no stake in the answer, the one who has not seen the numbers yet.

The full argument on spending paid budget responsibly is in how to run paid ads without burning the budget.

See the 90-day program — 13 weeks, real briefs, work that gets refused when it isn't good enough, and an independent assessment at Day 90 graded by someone who never taught you. Not everyone passes.

Written by Harry Aloysius. He runs Kreative Clan and teaches at School of Kreate.

FAQs

What should you do before you launch a campaign?

Before you launch a campaign, write a short setup note defining what the campaign is trying to achieve, the specific outcome that counts as success, the target cost, the stopping rule, the review date, what you are testing, the destination page, and who makes the final decision.

Why should you define campaign success before launch?

Defining success before launch prevents the campaign from being judged according to whatever happens after the money starts moving. A target and clear outcome give you a fixed standard for deciding whether the campaign worked.

What is a stopping rule in a paid advertising campaign?

A stopping rule defines when an individual ad should be stopped based on a pre-agreed spend threshold and performance metric. For example, an ad might be turned off if it spends twice the target cost per enquiry without producing one.

How long should you wait before reviewing a new campaign?

Set the review date before launch and avoid making changes simply because early performance looks disappointing. The setup note should also define what counts as an emergency, such as a broken link, incorrect creative, or obvious overspend.